Epstein-Barr virus-associated post-transplant lymphoproliferative disease (EBV-PTL) is one of the few human malignancies that is unequivocally related to immune dysfunction. The incidence of fatal EBV-PTL depends largely on the degree of post-transplant immunosuppression and may exceed 20% in recipients of bone marrow grafts from unrelated donors or from HLA antigen mismatched family members. The tumor cells in this disease have a normal karyotype and are similar in cell surface phenotype and virus gene expression to the immortalized lymphoblastoid cell lines (LCLs) that are established after infection of B lymphocytes with EBV in vitro. LCLs are exquisitely sensitive to killing by HLA-restricted EBV-specific cytotoxic T lymphocytes (CTLs), cells that cannot be detected in marrow recipients in the post-transplant period. We hypothesize first, that EBV-LPD should be prevented by reconstitution of susceptible patients with donor-derived EBV-specific CTLs and second that if the CTLs are marked by transduction with a neomycin resistance gene, their fate and function after infusion can be monitored. These hypotheses will be tested in the following three specific aims: 1. To establish donor-derived, gene-marked, EBV-specific cytotoxic T cell lines which can effectively be monitored in vivo with regard to function, longevity and function. 2. To determine the safety of infusions of BMT donor derived EBV-specific cytotoxic T lymphocytes marked with the neomycin-resistance gene in patients transplanted with marrow from a matched unrelated or mismatched family member for leukemia. 3. To evaluation survival and immunological efficacy of EBV-specific CTL in a bone marrow transplant population with high incidence of post- transplant lymphoproliferative disorders.